FTneuralCBF is a toolbox for designing NN-based fault-tolerant control as well as fault-detection and isolation (FDI) mechanisms. For more details, please refer to the paper: K. Garg, C. Dawson, K. Xu, M. Ornik and C. Fan, "Model-free Neural Fault Detection and Isolation for Safe Control," in IEEE Control Systems Letters, doi: 10.1109/LCSYS.2023.3302768
git clone https://github.com/kunalgarg42/NeuralFaultDetector.git
conda create --name [CONDA ENV NAME] python=3.9
conda activate [CONDA ENV NAME]
pip install -r . requirements.txt
In order to setup the learning, first create a python file for your control-affine systems in the dynamics folder (using control_affine_system_new.py) as the base file.
Then, for CBF + u learning, create a train file following the setup of Crazyflie_train_new file. For training FDI, use CF_train_Gamma_Output file.
Finally, for testing the performance of FDI, you can use CF_test_Gamma_compare_NNs_output.py as the base file.